Dynamic and Real-Time Object Detection Based on Deep Learning for Home Service Robots.

Home service robots operating indoors, such as inside houses and offices, require the real-time and accurate identification and location of target objects to perform service tasks efficiently. However, images captured by visual sensors while in motion states usually contain varying degrees of blurriness, presenting a significant challenge for object detection. In particular, daily life scenes contain small objects like fruits and tableware, which are often occluded, further complicating object recognition and positioning. A dynamic and real-time object detection algorithm is proposed for home service robots. This is composed of an image deblurring algorithm and an object detection algorithm. To improve the clarity of motion-blurred images, the DA-Multi-DCGAN algorithm is proposed. It comprises an embedded dynamic adjustment mechanism and a multimodal multiscale fusion structure based on robot motion and surrounding environmental information, enabling the deblurring processing of images that are captured under different motion states. Compared with DeblurGAN, DA-Multi-DCGAN had a 5.07 improvement in Peak Signal-to-Noise Ratio (PSNR) and a 0.022 improvement in Structural Similarity (SSIM). An AT-LI-YOLO method is proposed for small and occluded object detection. Based on depthwise separable convolution, this method highlights key areas and integrates salient features by embedding the attention module in the AT-Resblock to improve the sensitivity and detection precision of small objects and partially occluded objects. It also employs a lightweight network unit Lightblock to reduce the network's parameters and computational complexity, which improves its computational efficiency. Compared with YOLOv3, the mean average precision (mAP) of AT-LI-YOLO increased by 3.19%, and the detection precision of small objects, such as apples and oranges and partially occluded objects, increased by 19.12% and 29.52%, respectively. Moreover, the model inference efficiency had a 7 ms reduction in processing time. Based on the typical home activities of older people and children, the dataset Grasp-17 was established for the training and testing of the proposed method. Using the TensorRT neural network inference engine of the developed service robot prototype, the proposed dynamic and real-time object detection algorithm required 29 ms, which meets the real-time requirement of smooth vision.

Design and Control of a Tendon-Driven Robotic Finger Based on Grasping Task Analysis.

To analyze the structural characteristics of a human hand, data collection gloves were worn for typical grasping tasks. The hand manipulation characteristics, finger end pressure, and finger joint bending angle were obtained via an experiment based on the Feix grasping spectrum. Twelve types of tendon rope transmission paths were designed under the N + 1 type tendon drive mode, and the motion performance of these 12 types of paths applied to tendon-driven fingers was evaluated based on the evaluation metric. The experiment shows that the designed tendon path (d) has a good control effect on the fluctuations of tendon tension (within 0.25 N), the tendon path (e) has the best control effect on the joint angle of the tendon-driven finger, and the tendon path (l) has the best effect on reducing the friction between the tendon and the pulley. The obtained tendon-driven finger motion performance model based on 12 types of tendon paths is a good reference value for subsequent tendon-driven finger structure design and control strategies.

A Novel CNN-BiLSTM Ensemble Model With Attention Mechanism for Sit-to-Stand Phase Identification Using Wearable Inertial Sensors.

Sit-to-stand transition phase identification is vital in the control of a wearable exoskeleton robot for assisting patients to stand stably. In this study, we aim to propose a method for segmenting and identifying the sit-to-stand phase using two inertial sensors. First, we defined the sit-to-stand transition into five phases, namely, the initial sitting phase, the flexion momentum phase, the momentum transfer phase, the extension phase, and the stable standing phase based on the preprocessed acceleration and angular velocity data. We then employed a threshold method to recognize the initial sitting and the stable standing phases. Finally, we designed a novel CNN-BiLSTM-Attention algorithm to identify the three transition phases, namely, the flexion momentum phase, the momentum transfer phase, and the extension phase. Fifteen subjects were recruited to perform sit-to-stand transition experiments under a specific paradigm. A combination of the acceleration and angular velocity data features for the sit-to-stand transition phase identification were validated for the model performance improvements. The integration of the CNN, Bi-LSTM, and Attention modules demonstrated the reasonableness of the proposed algorithms. The experimental results showed that the proposed CNN-BiLSTM-Attention algorithm achieved the highest average classification accuracy of 99.5% for all five phases when compared to both traditional machine learning algorithms and deep learning algorithms on our customized dataset (STS-PD). The proposed sit-to-stand phase recognition algorithm could serve as a foundation for the control of wearable exoskeletons and is important for the further development of intelligent wearable exoskeleton rehabilitation robots.

A Semi-Autonomous Hierarchical Control Framework for Prosthetic Hands Inspired by Dual Streams of Human.

The routine use of prosthetic hands significantly enhances amputees' daily lives, yet it often introduces cognitive load and reduces reaction speed. To address this issue, we introduce a wearable semi-autonomous hierarchical control framework tailored for amputees. Drawing inspiration from the visual processing stream in humans, a fully autonomous bionic controller is integrated into the prosthetic hand control system to offload cognitive burden, complemented by a Human-in-the-Loop (HIL) control method. In the ventral-stream phase, the controller integrates multi-modal information from the user's hand-eye coordination and biological instincts to analyze the user's movement intention and manipulate primitive switches in the variable domain of view. Transitioning to the dorsal-stream phase, precise force control is attained through the HIL control strategy, combining feedback from the prosthetic hand's sensors and the user's electromyographic (EMG) signals. The effectiveness of the proposed interface is demonstrated by the experimental results. Our approach presents a more effective method of interaction between a robotic control system and the human.

A High-Efficient Reinforcement Learning Approach for Dexterous Manipulation.

Robotic hands have the potential to perform complex tasks in unstructured environments owing to their bionic design, inspired by the most agile biological hand. However, the modeling, planning and control of dexterous hands remain unresolved, open challenges, resulting in the simple movements and relatively clumsy motions of current robotic end effectors. This paper proposed a dynamic model based on generative adversarial architecture to learn the state mode of the dexterous hand, reducing the model's prediction error in long spans. An adaptive trajectory planning kernel was also developed to generate High-Value Area Trajectory (HVAT) data according to the control task and dynamic model, with adaptive trajectory adjustment achieved by changing the Levenberg-Marquardt (LM) coefficient and the linear searching coefficient. Furthermore, an improved Soft Actor-Critic (SAC) algorithm is designed by combining maximum entropy value iteration and HVAT value iteration. An experimental platform and simulation program were built to verify the proposed method with two manipulating tasks. The experimental results indicate that the proposed dexterous hand reinforcement learning algorithm has better training efficiency and requires fewer training samples to achieve quite satisfactory learning and control performance.

Co-optimization of robotic design and skill inspired by human hand evolution.

During evolution of the human hand, evolutionary morphology has been closely related to behavior in complicated environments. Numerous researchers have revealed that learned skills have affected hand evolution. Inspired by this phenomenon, a co-optimization approach for underactuated hands is proposed that takes grasping skills and structural parameters into consideration. In our proposal, hand design, especially the underactuated mechanism, can be parameterized and shared with all the local agents. These mechanical parameters can be updated globally by the independent agents. In addition, we also train human-like 'feeling' of grasping: grasping stability is estimated in advance before the object drops, which can speed up grasping training. In this paper, our method is instantiated to address the optimization problem for the torsion spring mechanical parameters of an underactuated robotic hand with multi-actuators, and then the optimized results are transferred to the actual physical robotic hand to test the improvement of grasping. This collaborative evolution process leverages the dexterity of the multi-actuators and the adaptivity of the underactuated mechanism.

An Enhanced Positional Error Compensation Method for Rock Drilling Robots Based on LightGBM and RBFN.

Rock drilling robots are able to greatly reduce labor intensity and improve efficiency and quality in tunnel construction. However, due to the characteristics of the heavy load, large span, and multi-joints of the robot manipulator, the errors are diverse and non-linear, which pose challenges to the intelligent and high-precision control of the robot manipulator. In order to enhance the control accuracy, a hybrid positional error compensation method based on Radial Basis Function Network (RBFN) and Light Gradient Boosting Decision Tree (LightGBM) is proposed for the rock drilling robot. Firstly, the kinematics model of the robotic manipulator is established by applying MDH. Then a parallel difference algorithm is designed to modify the kinematics parameters to compensate for the geometric error. Afterward, non-geometric errors are analyzed and compensated by applying RBFN and lightGBM including features and kinematics model. Finally, the experiments of the error compensation by combing combining the geometric and non-geometric errors verify the performance of the proposed method.

Pneumatically Actuated Soft Gripper with Bistable Structures.

This study presents the design and test of a novel self-adaptive soft gripper, integrating pneumatic actuators and bistable carbon-fiber reinforced polymer laminates. The morphology was designed using the distinct structural characteristics of bistable structures; and the stable gripping configuration of the gripper was maintained through the bistability without continuous pressure application. The sufficient compliance of bistable structures makes the gripper versatile and adaptable to gripping deformable objects. First, a pneumatic-actuated method was introduced to achieve the reversible shape transition of the bistable structure. Next, three arrangement methods for actuators were analyzed with respect to the bistable transition and curvature, where it was found that the cross-arrangement is optimal. The effects of pneumatic actuators with different geometrical parameters on the response times are discussed, and the results show that the bistable structure can achieve shape transition within milliseconds under low pressure. Furthermore, the numerical and experimental results show good agreement between critical pressures and out-of-plane deformation. Furthermore, the shape retention function of the soft gripper was studied by using it to grasp objects of various sizes even when the pressure was reduced to the initial state. The bistable laminates exhibit sufficient compliance, and the deformed laminates can automatically accommodate the deformation of objects. The relationship between the weight and size of available gripping objects was studied; functional tests confirmed that the proposed soft gripper is versatile and adaptable for gripping objects of various shapes, sizes, and weights. This gripper has immense potential to reduce energy consumption in vacuum environments such as underwater and space.

Psychometric properties of the Chinese version of the preoperative assessment of readiness tool among surgical patients.

Background: The evaluation of the surgical readiness of patients plays an important role in clinical care. Preoperative readiness assessment is needed to identify the inadequacy among surgical patients, which provides guide for interventions to improve patients' preoperative readiness. However, there is a paucity of high-level, quality tool that evaluate surgical readiness of patients in China. The purpose of this study is to translate the Preoperative Assessment of Readiness Tool (PART) into Chinese and determine the reliability and validity of the Chinese version in the population of surgical patients. Methods: Using a standard translation-backward method, the original English version of PART was translated into Chinese. A convenient sampling of 210 surgical patients was recruited from 6 hospitals in Zhejiang Province to test the psychometric properties of this scale including internal consistency, split-half reliability, content validity, structure validity, and floor/ceiling effect. Results: A total of 194 patients (92%) completed questionnaires. The Chinese version of PART achieved Cronbach's alphas 0.948 and McDonald's omega coefficient 0.947, respectively, for the full scale. The estimated odd-even split-half reliability was 0.959. The scale-level content validity index was 0.867, and the items content validity index ranged from 0.83 to 1.0.The output of confirmatory factor analysis (CFA) revealed a two-factor model (χ2 = 510.96; df = 86; p < 0.001; root mean square error approximation = 0.08) with no floor/ceiling effect. Conclusion: The Chinese version of PART demonstrated acceptable reliability and validity among surgical patients. It can be used to evaluate patients' preoperative preparation and help health professionals provide proper preoperative support.

Accuracy of self-perceived risk of falls among hospitalised adults in China: an observational study.

OBJECTIVE: To evaluate the accuracy of self-perceived risk of falls in hospitalised adults and explore factors associated with the differences. DESIGN: Cross-sectional study. SETTING: We conducted the study in two tertiary general hospitals located in Zhejiang province and Shandong province in China. PARTICIPANTS: 339 patients were recruited using convenient sampling. The majority of them were men (54%), aged 61-70 (40.1%) and had received secondary school education or lower (82%). OUTCOME MEASURES: The Fall Risk Perception Questionnaire and the Morse Fall Scale (MFS) were used to measure patients' self-perceived risk of falls and nurses' assessment. Other risk factors of falls were assessed to identify the determinants of disparities. RESULTS: Most patients (74.6%) had a high risk of falls according to MFS. Only 61.9% of the patients' perceived risk matched with the assessment of nurses. Nearly one-third (27.5%) underestimated their fall risk, while the remaining (10.6%) overestimated. Multivariable logistic regression analyses revealed that older age, lower number of comorbidities, not having fear of falling and emergency department were the significant factors associated with underestimated risk of falls (p<0.05). Besides, endocrine department and having fall-related injuries were significantly associated with overestimated risk of falls (p<0.05). CONCLUSION: Hospitalised patients were proven to be poor at recognising their risk of falls. Measurement of patients' self-perceived and health professionals' assessment of fall risk should be conducted to evaluate the disparity. This study provides a solid foundation to raise medical staff's awareness of the targeted population, identify the underlying factors and implement tailored fall prevention strategies and education.

Machine learning predicts ecological risks of nanoparticles to soil microbial communities.

With the rapid development of nanotechnology in agriculture, there is increasing urgency to assess the impacts of nanoparticles (NPs) on the soil environment. This study merged raw high-throughput sequencing (HTS) data sets generated from 365 soil samples to reveal the potential ecological effects of NPs on soil microbial community by means of metadata analysis and machine learning methods. Metadata analysis showed that treatment with nanoparticles did not have a significant impact on the alpha diversity of the microbial community, but significantly altered the beta diversity. Unfortunately, the abundance of several beneficial bacteria, such as Dyella, Methylophilus, Streptomyces, which promote the growth of plants, and improve pathogenic resistance, was reduced under the addition of synthetic nanoparticles. Furthermore, metadata demonstrated that nanoparticles treatment weakened the biosynthesis ability of cofactors, carriers, and vitamins, and enhanced the degradation ability of aromatic compounds, amino acids, etc. This is unfavorable for the performance of soil functions. Besides the soil heterogeneity, machine learning uncovered that a) the exposure time of nanoparticles was the most important factor to reshape the soil microbial community, and b) long-term exposure decreased the diversity of microbial community and the abundance of beneficial bacteria. This study is the first to use a machine learning model and metadata analysis to investigate the relationship between the properties of nanoparticles and the hazards to the soil microbial community from a macro perspective. This guides the rational use of nanoparticles for which the impacts on soil microbiota are minimized.

Corrections to "Academic Review and Perspectives on Robotic Exoskeletons".

We have noticed some errors in the above-titled paper (DOI: 10.1109/TNSRE.2019.2944655) [1].

Academic Review and Perspectives on Robotic Exoskeletons.

Since the first robotic exoskeleton was developed in 1960, this research field has attracted much interest from both the academic and industrial communities resulting in scientific publications, prototype developments and commercialized products. In this article, to document the progress in and current status of this field, we performed a bibliometric analysis. This analysis evaluated the publications in the field of robotic exoskeletons from 1990 to July 2019 that were retrieved from the Science Citation Index Expanded database. The bibliometric analyses were presented in terms of author keywords, year, country, institution, journal, author, and the citation. Results show that currently the United States has taken the leading position in this field and has built the largest collaborative network with other countries. The Massachusetts Institute of Technology (MIT) made the greatest contribution to the field of robotic exoskeleton investigations in terms of the number of publications, average citations per publication and the h-index. In addition, the Journal of NeuroEngineering and Rehabilitation ranks first among the top 20 academic journals in terms of the number of publications related to robotic exoskeletons during the period investigated. Author keyword analysis indicates that most research has focused on rehabilitation robotics. Biomedical engineering, rehabilitation and the neurosciences are the most common disciplines conducting research in this area according to the Web of Science (WoS). Our study comprehensively assesses the current research status and collaboration network of robotic exoskeletons, thus helping researchers steer their projects or locate potential collaborators.

Soft Robotics: Academic Insights and Perspectives Through Bibliometric Analysis.

Soft robotics is of growing interest in the robot community as well as in public media, and there is an increase in the quality and quantity of publications related to this topic. To formally elaborate this growth, we have used a bibliometric analysis to evaluate the publications in the field from 1990 to 2017 based on the Science Citation Index Expanded database. We present a detailed overview and discussion based on keywords, citation, h-index, year, journal, institution, country, author, and review articles. The results show that the United States takes the leading position in this research field, followed by China and Italy. Harvard University has the most publications, high average number of citations per publication and the highest h-index. IEEE Transactions on Robotics ranks first among the top 20 academic journals publishing articles related to this field, whereas Soft Robotics holds the top position in journals categorized with "ROBOTICS." Actuator, fabrication, control, material, sensing, simulation, bionics, stiffness, modeling, power, motion, and application are the hot topics of soft robotics. Smart materials, bionics, morphological computation, and embodiment control are expected to contribute to this field in the future. Application and commercialization appear to be the initial driving force and final goal for soft robots.

Atrazine affects the circadian rhythm of Microcystis aeruginosa.

This study provides original data regarding the effects of atrazine (Atr) on the circadian rhythm of the cyanobacterium Microcystis aeruginosa. The results reveal that the circadian rhythms of the central circadian oscillator genes reached their peaks from 1 to 2.5 h after the light was switched on, and the circadian rhythms of physiologically related genes were highly synchronized with the central circadian oscillator genes. These circadian rhythms were consistent with cell growth at the physiological level. The circadian rhythms of the central circadian oscillator genes were altered, and their peaks disappeared or were delayed by the Atr treatment. Therefore, the rhythms of the physiologically related genes in this study also changed to synchronize the new circadian rhythms. And the physiological parameters were tightly correlated with the gene circadian rhythm in the Atr treatment, suggesting that Atr affects M. aeruginosa growth by possibly altering the circadian expression patterns of the clock. Furthermore, this influence is related to the exposure time point of Atr. Thus, chemicals treated in the suitable exposure time point can exert their fullest effects against cell growth.